 Computers keep changing the world, but their power and safety is limited by their rigid design. The T2TILE project works for bigger and safer computing using living systems principles. Follow our progress here on T Tuesday updates. I'm Dave Ackley. This is the eighth update in the T2TILE project. Let's get into it. Our goal from last week was to release ULAM MFM and SPLAT, ULAM 4, the new version of the programming language, MFM, the underlying, the movable feast machine, the underlying engine for performing indefinitely scalable computations, and SPLAT, the spatial programming language based on ASCII text that's laid on top of ULAM to allow expressing spatial relationships in a compact fashion. We have the sort of mixed happy sad face. We'll talk about that in a minute. The ULAM 4 and the new version of MFM, one of the primary things that they were going to provide was providing support for brick wall geometry for the T2TILEs, which is the project. So, I mean, so here's a T2TILE. We saw it before. And the point is, well, I don't know another one, but here's just a case front, is that when they meet up, they're supposed to meet up like bricks staggered in different rows, but when they're in the same row, they can line up square. And, you know, so actually, you know, these things are designed so that if you look at the case, there's kind of little knobs on it so that if you try to do it checkerboard style, it doesn't really fit. And you can't get the connectors quite so close to each other in the spirit of robust computation. Our goal for next week is to take one of the new T2 PCB boards, which we've been waiting for for a couple of weeks now, and actually get it built, get it running, and test the light sensor, which was one of the main things in addition to changing the mounting holes that we were doing this revision for. Right now, we're going to talk about little bits of news and how it went with the attempt to release Oolong 4, and then deal with one question that came in last week. All right, and I'm going to finish in 20 minutes, just you watch. All right. So, oh yes. So look at this. How cool is this? I got my Carpe Event window t-shirt came in from MFM Rock store. It's very nice. This is the Gildan one. It's pretty smooth. And it's nice. It's great. I don't know if I'm really, you know, feel like, you know, I can't get too excited about these things because we're all at the very beginning and we're trying to figure out where I'm trying to figure out ways to make this project move forward, figure out ways to move the ball down the field sort of any way I can. But, and, you know, I don't want to take, you know, pay off. I don't want to pretend that we've gotten further than we have a tremendous amount to do. A lot of ways that this could screw up a lot of ways that even if it doesn't screw up on its own, it could get ignored by society and so on. But still, t-shirts. All right. So that's that. We also, as I mentioned, we have the new version of the PCB board, PCB board PCB. It's got the standard credo on the back and it's got the rounded sort of much more traditional mounting holes on the sides. I got to get some brass fittings. I got to dig them up to get it in there. And it's got a little teeny place over in the corner here. You can't really see it, which is where the light sensor would go. So if we have two tiles next to each other, the light sensor would be looking up in the gap between them. We'll see how it works. So that's all coming up. Oh, and just in the spirit of, you know, show and tell and learning a little bit about how all of this hardware stuff works for folks that don't know, since I really don't know where's the soap up here. I also got the stencil that goes with this. Did I? Oh, I taped it down again. That's why it's not coming up. And the idea is, is that if you have a stencil, you instead of trying to put down tape by a tape, put down solder by hand in each teeny little place on the board, you can instead, there we go, put the board underneath the stencil. Here it is. I don't know if you can see that. If we get the lights a little better. All through here, there's all these little teeny holes that you put the board on one side, you put the stencil on top, and you get this paste solder and you squeegee it through so that you get little blobs of solder put on all the things automatically. It's easier said than done. I've tried it a couple of times before with mixed success. So one of the things that hopefully we'll talk about next week is hand soldering versus reflow soldering. It's called where you use a stencil and paste. Okay. That's it for the news. Let's talk about Oolong 4, getting Oolong 4 out. Now, I thought the sort of teacher on me makes it hard for me to resist trying to take a little bit of a bigger perspective on whatever the task of the day is, because that's just what I do. Living computation, everything's about computation. And the whole idea of distributions is something that's very funny to me, because a long time ago in the 90s, in the mid 90s, when I first got to New Mexico and I was just starting as a professor, as a college professor, I knew about Linux. I was using Linux and I started hearing about these distros and it was like, what the heck is a distro? And it seemed to me at the time that these people that were making distros were sort of con artists. They were sort of cheaters. They were taking somebody else's real work, writing the operating system. They're taking Linux and they were just kind of duding it up by putting in a few other little programs to go with it and then giving it a whole new name, calling it a distro instead of an operating system. And I didn't see it at the time, but eventually, as I was just today, as I was making up these slides, this was the announcement in 1993 of the Debian project from Ian Murdoch and what the heck was a distro. And the reason I bring it up now is because how did it go? Well, you know, on the one hand, distros are such a huge fundamental thing now. Distributions, combinations of programs that play well together that build on top of the operating system and they've gotten so much structure and so much design that I can have my own little piece, my personal package archive that is where Ulam and MFM and now Splat as well are all located. And that's because I was completely wrong when I was thinking, oh, all the work is in Linux, all the work is in the operating system and just, you know, getting a magnetic tape and putting a bunch of other programs next to it, who cares? In fact, of course, this is making the transition to the next level of spatial complexity. Yes, you know, the operating system does a tremendous amount of stuff. Does that mean the job is done? No, it's not. And so you're naturally going to have next levels of complexity and each one of those is going to look like, you know, standing on the shoulders of giants if you're feeling kind about it or like kind of ripping off the work of the people before if you're feeling curmudgeonly about it. For example, these days, Android, everybody knows about Android, but the typical person who knows about Android doesn't know that it's really Linux and just got another layer of complexity put on it to deal with phone hardware and so on and so forth. So, distros are a legitimate next thing. We have to keep that in mind when we're dealing in the movable feast and it's flat out when we're writing these tiny little programs for fork bombs and building boxes and so on to not think that that's going to be it. That there's going to be layers upon layers upon layers of now they're going to be spatially bigger as they go. And it's going to be just like operating system to distro and so on. All right, so how did it go? Well, of course, the way it went was, you know, rejected error crash, you know, the way it always goes. The fact that I had forgotten how to hadn't done this for over a year didn't help the fact that a bunch of the tools had rotted out underneath me as well as my own environment and I had changed machines and didn't have the right keys in the right places. All right, I got it sorted out. And, you know, and well, this is the other thing, you know, the Ubuntu distribution they have this incredible giant build farm, hundreds of machines that are just there continually rebuilding packages that have been updated and need to be made into binaries so that they can be distributed. And so I uploaded my stuff and I thought, you know, last time I did this it took like 10 minutes or so 15 minutes before I got into get my thing built and I was like, well, geez, today, you know, two hours and 40 minutes that's after my job started and then it failed at the end. So, you know, basically there's something going on with the 32 bit Intel architecture. Ulam is not building on the Intel 32 bit architecture. I don't know why I'll have to debug that going forward and especially because for one thing, you know, the T2 tile is a 32 bit architecture. It's not Intel, but it is 32 bits. So we'll see what's going on there. But the AMD 64, which, you know, most machines these days are 64 bit machines the machine that I'm recording this on 64 bit machine. The packages are there. Ulam, MFM, Splat, all there, all installable. Do I have a, I guess I don't have a thing, but I have the actual thing. Oh, here's my sound guy. So now I don't have to do dot slash bin, whatever. I can just do MFZ run and it's all installed. So MFZ run DLA and look at this now out of the box it defaults to staggered, it defaults to brick wall geometry, but the stuff still works. This is the diffusion limited aggregation demo where you start with one seed and those just res all bouncing around and whenever res finds itself, well, whenever the DLA finds itself next to a res, it freezes it and turns it into a DLA and you get this kind of like sort of frost like pattern growing out from the original seed. Anyway, it's packages, you just get the update. It's great. It's great. A lot of progress. All right, well, we're not going to take time to look at that anymore. Even though it's great, you should do it yourself. All right. So bottom line, the 64 bit architectures are in the repose done. The 32 bit architectures still need work. The arm one, which is actually what the T2 tiles are using also crashed. Had a bug that's now been fixed, but I didn't have time because I was making the video to go about uploading it again. So that's going to take a while. So that's where we're at. All right. Hey, Andy X. Thank you so much for participating. You know, I see you've been kind of poking around the movable feast and stuff for a few years. One way or another, I see your name. For not as enterprising as Andrew or Spencer. Yeah, really, folks that would happy be happy doing some kind of relatively unskilled labor testing, whatever, purchasing a few tiles flash and run the system. You know, that's actually not a small thing. As I mentioned to somebody in a comment, these tiles are going to be way expensive. When people get in their mind, they expect them to be like 3040 dollars like a raspberry pie. They're going to be $100 plus at my cost. So I doubt that that's not a small thing. If, hey, Andy X, if you happen to be like super rich and you want to come on as a patron and buy a chunk of tiles to help grow the machine. If anybody else does, let's talk. But I wouldn't expect that of anybody feature request. Nice to have, you know, yes, there's all these things that make sense to be wanting to have. But I don't have time to do them all. And I don't actually really have the skills to do it very well. I'm kind of, you know, previous generation as far as digital native and so forth. I mean, I'm pretty fluent, but I'm still immigrant to digital nativity. And so and Andrew brought it up, for example, do we need a discord server? Is that a good thing? What's that for? Do we need a straight up wiki that anybody can contribute? There is the robust wiki at robust.cs. But you have to have an account on UNMCS, which not many people in the universe though. In order to work on that, as much as anything, I need suggestions from y'all from from people about how to actually do this. Perhaps even some hobnobbing amongst yourselves to figure out what would be a good idea? What would collectively help you? And I say, yes, I want that to happen. Feature guest and at least a place to collect all that information, you know, absolutely. And so this is what I was saying. Okay, if we're not talking about speaking. Oh, dear, did I lose? Oh, yeah. I had a bunch more things that didn't make it in. Well, it's too bad. So what I meant by so if we're saying not coding and that's fine. Everybody can write, you know, everybody that does a video at the end, they say, you know, please share, please subscribe and everything. And I've tried not to do that because to me it sounds so kind of forced and everything. And even though I really understand now why people say it, I still have kind of resisting it. But there are other ways. I mean, one of the things that we need to do is get more buzz in more different circles that, you know, there is this other perspective on traditional computation that a lot of people don't seem to know about. And we need to increase awareness of there is this alternative, even though it's very primitive and it's going to take a lot of work and so forth. So the examples that I had up here and then managed to lose. Don Hopkins was writing comments on hacker news. I didn't even know that hacker news really existed. It seems like it's kind of a slash dot next generation or something like that. And he had this wonderful thing where he someone was talking about 2d is so limiting. And he actually kind of transcribed a chunk of the answer to a question from my retirement talk last May about 3d versus 2d and put it up there. And that is fantastic. And to me, that's an example of sort of sharing and building awareness that isn't just, you know, subscribe and retweet like that. You know, it's actually, you know, the organic stuff where people are using their own interpretation, just like Andrew was doing and so forth to get it out there. So I want to, you know, if anybody can think of places to speak about it to talk into to build awareness, that would be super. And speaking up, the general point is, if this is all about not being a serial determinism, it's all about being distributed, then ultimately I can't be the one to know everybody has to be saying, well, we need this, we need this, well, you know, that'd be fine, but it's not going to go very far for that. So any way that you can find to speak up about this stuff is good. And certainly even in the simplest way of making comments and asking questions and so forth, that helps me tremendously. You know, I started with a little voice in the wilderness for quite some period of time. And I really, really, really enjoy even if, you know, well, if the questions get kind of repetitive, then that goes to writing, right? We need to come up with, where should a fact go? We need a fact. And we ought to be able to collect questions. We ought to be able to collect draft answers. And I am much, much, much better at rewriting than I am at writing. So if, you know, text can be started up and collected up in some kind of rough form, then I would enjoy sort of swooping in and saying, well, that's not exactly how it works. But, you know, like that and trying to make it a little bit closer to truth as well as a little bit closer to what people can understand. So, hey, Andy X, thank you for all you've done so far. If anybody is interested in trying to help the mission in any way they can, even if it's not buying hardware or debugging hardware or buying tiles, then, you know, talk to people about it, spread it out in the world, and keep speaking to me as much as you can, because it really does help. So that's what I can think of. Okay. And so we're under 20 minutes. I'm going to stop next week. We're going to see a new board running and the light sensor working. Hopefully I can tell you I've already done a little bit of work and there are some little hitches. So we'll find out about that work next week. The next update will be out in a week. Thanks so much for being here. Thanks so much for watching. I'll see you in 18 and a half minutes.